Testosterone

What it actually measures

Most of the testosterone in your blood is not doing anything, and that is the fact this whole page turns on.

It circulates bound to carrier proteins. The majority binds tightly to sex hormone binding globulin, SHBG, and hormone bound that way cannot enter a cell and act. A smaller portion binds loosely to albumin, and a small fraction circulates entirely free.

So there are three numbers. Total testosterone is everything. Free testosterone is the unbound fraction that acts directly. Bioavailable is free plus the loosely albumin-bound portion, which is also reasonably accessible.

That matters because anything shifting your SHBG changes your total without changing what your tissues actually receive.

Thyroid disease shifts SHBG. Insulin resistance lowers it. Ageing raises it. Oral oestrogen raises it substantially, which is why hormonal contraceptives reduce free testosterone even when the total looks unremarkable.

Two people with the same total testosterone can therefore be in genuinely different situations, and only the fractions tell you which.

What moves the number before anything is wrong

Before interpreting any result, two things need accounting for, and both are commonly ignored.

The first is timing. Testosterone follows a strong daily rhythm, peaking in the early morning and falling through the day. The convention is a morning draw, generally before about 10am, and an afternoon result in someone with a perfectly normal morning level can look low.

If you are comparing against a previous result, drawing at a similar time removes a variable that otherwise looks like a change in you.

The second is sleep, and the effect is larger than its reputation.

A 2021 randomized double-blind in-laboratory crossover study in the Journal of Clinical Endocrinology and Metabolism opened by stating the established position: sleep loss in men increases cortisol and decreases testosterone.

The study then put 34 healthy young men through four nights of four-hour sleep twice, once with cortisol and testosterone clamped and once with placebo, and found that fixing those hormones cut the resulting insulin resistance by about half.

The relevant point here is what the study took as its premise. Short sleep lowers testosterone, reliably enough to build a trial design around.

Which means testing during a stretch of poor sleep risks measuring your schedule rather than your baseline, and it is worth sorting the sleep for a few weeks first where you can.

Third, there is a reference range problem that is genuinely a laboratory issue rather than a biological one. A 2017 study in the Journal of Clinical Endocrinology and Metabolism cross-calibrated assays across four cohorts totalling 9,054 community-dwelling men against a CDC reference method, and found a substantial share of the variation between cohorts was assay artefact rather than true biological difference.

It produced harmonised ranges. In healthy non-obese men aged 19 to 39, the 2.5th to 97.5th percentiles ran from 264 to 916 ng/dL, with a median of 531.

That is useful for two reasons. It gives a defensible range, and it explains why the same man tested at two labs can get results that appear to disagree.

Fourth, and worth saying plainly: the relationship between testosterone level and symptoms is looser than the advertising implies. Plenty of men in the lower part of the range have normal libido and energy, and plenty in the middle do not.

What moves it

Down: ageing, gradually and normally. Short sleep, measurably. Excess visceral fat and insulin resistance. Significant illness or surgery. Some medications including opioids, corticosteroids and certain others. Overtraining alongside under-eating. Pituitary problems, which lower the signal rather than the gland. And in women, hormonal contraceptives, which lower the free fraction by raising SHBG.

Up: prescribed testosterone therapy, which is a medical decision with real trade-offs including fertility. Improving sleep and body composition, which move it modestly and genuinely. Resistance training. In women, polycystic ovary syndrome raises it, alongside adrenal sources, which is why it appears in that workup.

The honest hedge

I am not a doctor and I am not diagnosing anyone, and this is one of the more heavily marketed numbers in the entire directory.

Testosterone therapy is a genuine medical decision with real trade-offs, including effects on fertility, and it belongs with a physician who investigates why a level is low rather than a clinic that advertises treating the number.

A good clinician reads levels alongside symptoms rather than one from the other, precisely because the correlation between them is looser than the marketing suggests.

For women, this marker is under-measured rather than over-measured, which is the opposite problem. Testosterone matters for libido, energy and muscle in women, and someone can go through an entire workup without anyone checking the hormone most directly tied to what they came in about. Because levels are lower, mass spectrometry is the more appropriate method than immunoassay.

The practical order I would suggest: fix sleep, check thyroid and iron, then measure, rather than measuring first and treating a number produced by a bad month.

And a low total with a normal free fraction is usually an SHBG story rather than a testosterone one.

So. What to actually do.

Tonight, at no cost. Set a bedtime that gives you seven and a half hours. Given that short sleep lowers this hormone reliably enough to build a study design around, that is the highest return free change here.

Draw it in the morning, before about 10am. Testosterone peaks in the early hours and falls through the day, so an afternoon draw can produce a low number in someone entirely normal.

Order the fractions, not just the total. Total, free and bioavailable testosterone with SHBG, because the carrier determines what your tissues actually receive.

Rule out the overlapping causes. A thyroid panel and ferritin with hs-CRP, since both produce symptoms that look like this and both are cheaper to fix.

Check your medication list. Opioids, corticosteroids and several others lower it, and that is a real explanation.

If it is genuinely low, see a physician rather than a clinic. The question of why matters, and therapy has real trade-offs including fertility.

For women, ask for it explicitly. It is measured far less often than it should be, and mass spectrometry is the appropriate method at lower concentrations.

You are not a single number on a marketing page. You are a hormone, a carrier protein that decides how much of it counts, and a sleep debt that nobody put on the requisition.

Your body is not broken. It is blocked. And often the block is upstream of the hormone entirely, in the hours you did not sleep and the thyroid nobody measured properly.

Go to bed earlier, then draw it in the morning.